Practis Labs / Applied Voice AI

Better conversations start with better Voice AI.

Behind every practice conversation is a human moment worth getting right.

Practis Labs is the applied Voice AI research and development group at Practis. We work on customer simulation, coaching and evaluation to help people prepare for real customer conversations.

Explore our work See the Practis platform

Our focus

Listen.
Respond.
Make room to learn.

Voice is the medium.
Human skill is the purpose.

Research with somewhere to go

Our work starts with the next conversation a person needs to have. The question they didn’t expect. The pause they want to handle differently. The chance to come back better prepared.

What we work on

From a convincing voice
to a useful practice partner.

Customer simulation

A customer with a point of view.

A hesitant homeowner. A buyer with a budget. A question that changes the direction of the conversation. Our simulation work focuses on making an AI Customer a useful practice partner, with context to respond to and a role to hold.

Explore AI Roleplay

Grounded coaching

Advice that belongs in the conversation.

Good coaching starts with the situation in front of someone. Our applied systems work connects company knowledge, conversation context and feedback, so a person can work on a specific skill and try it again.

Explore Practice & Coaching

Performance evaluation

A score needs something behind it.

A completed session says someone practiced. Understanding how they performed takes a clear standard and evidence from the conversation. Our evaluation focus is on the criteria that make feedback useful to people and their coaches.

Read the PRACTIS™ Method

How we think about quality

Sounding natural
is one part of the work.

A simulation needs to be useful to the person practicing. These questions guide our applied systems and evaluation work.

Our focus is the systems around the conversation: how models, context and feedback work together.

Does the simulation hold together?

Persona consistency, relevant responses and conversational timing all shape a practice session. These are separate questions from whether a voice sounds convincing.

Is the guidance grounded?

Feedback should connect to the conversation and the company knowledge it draws on. Unsupported assertions and missing context belong in the evaluation, too.

Would a human coach agree?

Human review matters when evaluating feedback. Clear rubrics help surface disagreement, explain what a score means and identify where a system needs more work.

Does practice carry into the real conversation?

Simulation quality and a person’s skill are different things. Understanding improvement requires looking beyond completed sessions to what someone can do in a customer conversation.

Where the work meets the working day

Built for the next appointment.

Practis gives people a place to rehearse, get feedback and try again before a real customer is listening. See how the product brings practice and coaching together.

Introducing Practis

Practice before
the next customer.

A place to try. Feedback to work with.
Another chance to get it right.

Watch on Vimeo

Practis Introduction / A product overview.

Explore the platform Read the Method

The people behind the work

Different disciplines.
One conversation.

Voice AI, platform engineering and product come together here. Ryan Felton leads the technical direction, working with the team building the experience people use.